Printrbot with Gen7+Slic3r: Print Quality issues

Hi Everyone

I build a Printrbot from the files on Thingiverse, and run it from a Generation 7 version 1.4, with Teacup firmware and Pronterface.
You can read more details on my build here:
[www.itux.co.za]
[www.itux.co.za]
[www.itux.co.za]

Now the issue I am having is that my printed parts have the following issues:

1. Even though the calibration cubes print with perfect exterior sizes, my interior sizes of all holes printed in the flat plane come out too small. i.e. X-Y movement. Holes printed in the sides of objects, come out fine.
I have done a lot of checks and my X/Y/Z calibration is not out by more than 0.1 - 0.2mm
The only theory I have is that I am extruding too wide and therefore the object is actually about half a extrution too narrow, and the width of the extrusion makes up the difference? This would also explain why holes are too small, as the extrution actually expans into the blank space where the hole should be. This is the biggest problem as for example in the print below, the linear bearings and m8 nuts doesnt fit in the holes.

2. I have precisely measured my filament feed with the straw and marker method. The 100% solid cube calibration test is perfect. However, ever here and there layers my nozzle stops extruding for a few mm.s This is fine on infill, but it messes up the exterior shells of my prints: Samples of print quality: [www.itux.co.za]. I Print PLA at 185 degrees - works perfect for Bridges so don't want to heat much higher, and I would expect temp issues to be more of a problem with no extrution followed by a blob...

3. Suspect belt tension:
[www.itux.co.za] - case in point - the exterior has random little bumps or it seems like the layers didnt stack exactly on top of each other. Could this be belt tension, or other issues with the Printed XL pulleys? Would it be worthwhile to upgrade to a proper XL pulleys bough from say RS-Online?

4. Stringing... a little bit of stringing, so might need to tweak retraction in Slic3r? [www.itux.co.za]

General print quality: Everyone with a bit more experience, just looking at my print above, can you see anything else I might be doing wrong?

Print speeds? Too fast, too slow?

Here's a paste of my Slic3r profile:

acceleration = 0
bed_size = 200,200
bed_temperature = 0
bottom_layer_speed_ratio = 0.3
bridge_fan_speed = 100
bridge_flow_ratio = 1
bridge_speed = 60
complete_objects = 0
cooling = 0
disable_fan_first_layers = 1
duplicate = 1
duplicate_distance = 6
duplicate_grid = 1,1
end_gcode = M104 S0 ; turn off temperature\nG28 X0 ; home X axis\nM84 ; disable motors
extruder_clearance_height = 20
extruder_clearance_radius = 20
extrusion_axis = E
extrusion_multiplier = 1.1
extrusion_width_ratio = 1.67
fan_always_on = 0
fan_below_layer_time = 60
filament_diameter = 3
fill_angle = 45
fill_density = 0.8
fill_pattern = rectilinear
first_layer_bed_temperature = 0
first_layer_height_ratio = 1
first_layer_temperature = 190
g0 = 0
gcode_arcs = 0
gcode_comments = 0
gcode_flavor = teacup
infill_acceleration = 50
infill_every_layers = 1
infill_speed = 60
layer_gcode =
layer_height = 0.30
max_fan_speed = 100
min_fan_speed = 35
min_print_speed = 10
notes =
nozzle_diameter = 0.5
output_filename_format = [input_filename_base].gcode
perimeter_acceleration = 25
perimeter_speed = 30
perimeters = 3
post_process =
print_center = 75,75
randomize_start = 1
retract_before_travel = 2
retract_length = 2
retract_lift = 0
retract_restart_extra = 0
retract_speed = 30
rotate = 0
scale = 1
skirt_distance = 6
skirt_height = 1
skirts = 2
slowdown_below_layer_time = 15
small_perimeter_speed = 30
solid_fill_pattern = rectilinear
solid_infill_speed = 60
solid_layers = 2
start_gcode = \nG28
support_material = 0
support_material_tool = 0
temperature = 188
threads = 2
travel_speed = 130
use_relative_e_distances = 0
z_offset = 0

Hoping you could help me, as I need to print some parts for a few bud's

Peter

I printed a Belt tensioner for the X axis and tightened the Y axis belt again. cleaned up the random moves - i.e. vertical pilars go up in a straight line now, but I still have that rippled surface issue...

What could be causing that?

Peter

Most likely the rippling is due to backlash in the Z-Axis. I had the exact same phenomena with my MendelMax 1.5 and it took quite a bit of trouble shooting to find out it was the captive nuts within the X-ends were able to slide around. I'm not familiar with how the X-Ends are designed on the PrintrBot but I was able to fix it by adding a spring and a second nut to apply pressure on the threads on both + and - z movements. The annoying banding has completely disappeared from my prints immediately after this modification.

Regarding the extruder pause it could be a couple of things. The Pololu's for the extruder might be overheating and shutting down temporarily. Turn down the trim pot (counter clockwise) until the extruder stops turning and then turn it back up (clockwise) until it extrudes again. The other potential problem could be the hot end temp dropping below your safe extrude temperature. Not sure if Teacup has a safe extrude temp but watch your temps to see if there are any abnormalities in your temperatures. I use Repetier as my host software which has a nice built in live graph to monitor temps / power usage of all heaters.

Stringing - yeah probably will need some tweaking of the retract.

Size Calibration - I like to leave my extrusion width ratio set to 0, slicer then automatically defines the best ratio for the settings that I have called out.

Hope this helps. How do you like the PrintrBot so far? I almost jumped on the PrinterBot+ before he sold out of them.

--Edit-- After looking at the PrinterBot X-Ends you could easily add a spring and a second anti-backlash nut. One of the biggest issues I see with typical instructions for assembling most RepRaps is that they typically rely on gravity to move the z-axis down (obviously the positive pitch of the leadscrews or threaded rod move the z-axis up). My problem was that even with LM10UU linear bearings and precision smooth rods, gravity was not sufficient to allow the z-axis to drop back down after a positive z-axis move. Adding the backlash nut essentially locked the z-axis assembly to the threaded rods allowing for much finer control.

Edited 1 time(s). Last edit at 07/12/2012 10:58PM by archistrong.

I'm loving the Printrbot - though it is my first build. Played with BFB Rapman 3.1, then a Replicator (great, but too expensive) before building the Printrbot.

I found one other Printrbot owner: Same issue: [blog.lincomatic.com]
He says he dropped the speed on the perimeter to get rid of vibrations.

I added the second nut in the top of the Printrbot Z axis as suggested, it didn't make major differences, but a bit. So thanks for that. Thats just second nut, not spring yet. I cannot feel any backlash on the Z.

What did make a major difference though was setting the Width over extrusion back to 0! Seems the wobbly edge is caused by too much plastic squeezing out when its infilling, and that presses against the perimeter and pressed it outwards every time the nozzle turns around at the perimeter.

I still had holes too small though...

This morning I upgraded to Slicer 0.84 (26 June release) and I don't know whats been changes since .83, but my holes are a lot closer. On 0.83 I couldnt even sand the holes down enough to press fit the LM8UUs, even after changing the Width over Extrusion.

But with 0.84 the holes are about 0.1mm off, enough that a quick sand with 100 grit sandpaper meant the LM8uu press fit easily...

Thanks for your help. We can mark this one closed, though I'll reply once more if the slower perimeter makes a big difference.

Peter